Time Travel: 2014

Huang Jing felt that Lin Hui was implicitly conveying to Huang Jing that he asked overseas original teams to create more designs that fit the habits of Chinese users.

At the same time, we should try our best to take care of Chinese users in design.

Why didn't Lin Hui say it clearly in the customization mission before?

Huang Jing felt that Lin Hui just missed her acquaintance with F.FSG, the original team of online office software.

Therefore, Lin Hui did not put forward too many suggestions for modifying the plan in front of Huang Jing.

Being able to be so considerate of the emotions of his subordinates made Huang Jing feel more and more that following Lin Hui was the right choice.

Of course, Lin Hui didn't know Huang Jing's inner self-strategy.

Later, Lin Hui and Huang Jing didn't talk about working online.

Instead, we chatted about some news from American technology giants.

Although it is basically boring news such as gossip, not all information is gossip.

At least Lin Hui didn't get nothing.

From the follow-up conversation with Huang Jing, Lin Hui learned a very important piece of information from Huang Jing.

That is, Apple seems to be committed to pursuing a large data transaction totaling approximately US$200 million to US$300 million.

Huang Jing was a bit vague when describing this news.

It seemed as if he was afraid of accidentally trapping Lin Hui.

The information described by Huang Jing in the past was often conclusive.

It is rare to feel unconfident.

When it comes to this transaction, Huang Jing first said it was a data transaction and then later said it was not a data transaction.

It made Lin Hui a little confused.

Even the gossip Lin Hui attaches great importance to the corresponding value. After all, there are often no smoke without fire.

As for what the message Huang Jing said was, it was subject to further inquiries and multiple checks.

After some further deliberation, Lin Hui finally figured it out.

The so-called data transaction of two to three billion US dollars does point to data, but it is not a general type of data transaction.

The data acquisition Apple is pursuing this time is actually a rather special data transaction.

Because of the information obtained through various channels, Lin Hui feels that Pingcheng’s goals are actually:

——"Dark data".

With this plan, it can be seen that Pingchun seems to be building a plank road to cross Chencang secretly.

Dark data is sometimes called dust data.

Dark data or “dust data” is made up of all redundant, often forgotten data.

This data is collected by companies and organizations in the course of their activities but then not used.

Dark data is often unstructured, unlabeled, and unanalysed information.

Compared with the annotated data that Lin Hui ignored before.

Dark data has no sense of existence.

Dark data This type of data is almost ignored.

After all, this kind of data exists in the network and server, and it only takes up valuable space.

Generally speaking, there are three main types of dark data:

The first is traditional text-based data. This may include emails, logs and documents.

The second type is non-traditional data.

This includes untagged audio and video files, still images, and sound files.

The third type is depth data.

This includes information in the deep web that search engines cannot reach.

Most of this deep data is private and controlled by governments or private entities.

It includes data, medical records, legal records, financial information and organization-specific databases curated by academics, government agencies and local communities.

All the above data can be called dark data.

…

Dark data is more obscure than data in the traditional sense.

Although unlabeled data such as dark data cannot be used directly.

But the potential of this kind of thing cannot be denied.

Anyway, it cannot be said that this information is not important.

As for why Guozi is interested in such things.

Because collecting this kind of data has always not been considered data.

In fact, through in-depth cultivation, you can get results similar to those of traditional data.

Moreover, by using this kind of data, through some conceptual education, consumers can even form the impression that the company never gets involved in general data.

Wouldn't this be very useful in establishing a corporate image? ?

In short, it cannot be said that it is not attractive to companies that are both relevant and established.

Anyway, Lin Hui feels that starting with dark data is in line with the behavior of many technology giants.

Compare it to Lin Hui’s previous estimated price.

If you say tens of millions of dollars, you can buy tens of millions of bilingual annotated data.

It is conceivable that dark data worth two to three billion US dollars, such as what Apple is seeking, must be a huge amount of data.

A major difference between annotated data and dark data is that annotated data is structured data that has undergone certain processing.

To a large extent, dark data is unstructured or even “messy” data.

Structured data is generally data that has a fixed format and limited length.

For example, a filled-in form is structured data.

For example, "Nationality, flower grower, ethnicity: Han, gender: male, name: Zhang San, age:..."

This format is called structured data.

This type of data is easily stored in a database in a fixed format.

Semi-structured data is worth some data in XML or HTML format.

This type of data can be processed as structured data as needed, or plain text can be extracted and processed as unstructured data.

The so-called unstructured data: data with variable length and no fixed format.

For example, web pages and emails are sometimes very long; sometimes they are very short and disappear in a few sentences. This type of data is typical unstructured data.

For example, Word documents, voices, videos, and pictures are all unstructured data.

Semi-structured data and unstructured data are generally combined into one and are collectively referred to as "dark data".

This word is not defined by Lin Hui.

Compared with structured data such as annotated data, the value of dark data and annotated data is not the same.

The value of unit labeled data is often dozens or even hundreds of times that of unit dark data.

Even if two to three billion US dollars are exchanged for more expensive cross-lingual language annotation data, it can be exchanged for hundreds of millions of pieces.

Not to mention spending hundreds of millions of dollars in exchange for dark data?

It is conceivable that the dark data involved in the two to three billion US dollars is a considerable amount of dark data.

Lin Huina has a lot of information about his past life.

But there is absolutely no dark data that can satisfy Apple’s appetite.

Not to mention the bit of information about Lin Hui's previous life.

Even the scale of dark data owned by some domestic Internet companies that are powerful and powerful among domestic Internet giants may not be able to satisfy Apple's appetite.

In this case, if Lin Hui is interested in Pingcheng's huge acquisition, it seems that he can only collect dark data.

As for how to collect it?

This is a problem. There are many ways to collect dark data, but not all of them can be accessed directly.